Roadmaps
Community-curated paths to learn to code. Pick one and a squad forms around you — you finish it together, not alone.
Community-curated paths to learn to code. Pick one and a squad forms around you — you finish it together, not alone.
Multiply your output as a developer with AI: prompting for code, the Claude API, agentic coding with Claude Code, subagents and skills, and MCP servers in your workflow. Best if you already write code.
What AI is reliably good at in a codebase, where it fails, and how to verify its work.
Give models the context, constraints, and examples that produce working code.
Call the model programmatically: messages, streaming, and system prompts.
Let an agent read, edit, and test your codebase from the terminal.
Scale up with specialized subagents, reusable skills, and headless runs.
Wire your tools — browsers, databases, issue trackers — into your AI assistant.